Changelog — recent AI API latency measurements

Updated September 07, 2026

Most recent probe results. A diff of ranking changes accumulates here as history builds — this is itself high-intent, citable content ("did [provider] get slower this week?").

Time (UTC)	Provider	Region	Probe	Latency	Status
2026-09-07T02:15:09	hunyuan	sa-east	network	1657 ms	ok
2026-09-07T02:15:09	iflytek	ap-tokyo	network	235 ms	ok
2026-09-07T02:15:08	ernie	sa-east	network	407 ms	ok
2026-09-07T02:15:07	stepfun	sa-east	network	632 ms	ok
2026-09-07T02:15:07	hunyuan	ap-tokyo	network	1124 ms	ok
2026-09-07T02:15:07	yi-01ai	sa-east	network	357 ms	ok
2026-09-07T02:15:07	stepfun	ap-tokyo	network	737 ms	ok
2026-09-07T02:15:07	sensenova	sa-east	network	646 ms	ok
2026-09-07T02:15:07	iflytek	sa-east	network	644 ms	ok
2026-09-07T02:15:07	sensenova	ap-tokyo	network	461 ms	ok
2026-09-07T02:15:06	doubao	ap-tokyo	network	449 ms	ok
2026-09-07T02:15:06	nscale	ap-tokyo	network	496 ms	ok
2026-09-07T02:15:06	baichuan	ap-tokyo	network	176 ms	ok
2026-09-07T02:15:06	ernie	ap-tokyo	network	136 ms	ok
2026-09-07T02:15:06	yi-01ai	ap-tokyo	network	189 ms	ok
2026-09-07T02:15:06	baichuan	sa-east	network	346 ms	ok
2026-09-07T02:15:06	targon	ap-tokyo	network	329 ms	ok
2026-09-07T02:15:06	sarvam	ap-tokyo	network	240 ms	ok
2026-09-07T02:15:06	aleph-alpha	ap-tokyo	network	530 ms	ok
2026-09-07T02:15:06	hunyuan	us-central	network	1375 ms	ok